Originally posted as a tweetstorm
In large product companies, most product efforts fall into two categories. One is a battle of “inches” and then there are those which are battles of “miles”. A friend calls this “trench warfare” vs “inventing a new kind of weapon” which may be more violent but still accurate imagery.
The “inches” efforts almost always involves rapid iteration and feedback. You’ll often find lots of AB testing and a strong focus on metrics. The team will celebrate a +0.1% metrics win and will have a culture of stacking up dozens of those. At any given time, they’ll have dozens of these in flight. The PMs on these teams will take pride of just looking at graphs all day and hunting down tiny optimizations. You often find these teams in what used to be called “growth teams”. A better pattern to match is that you find them in product efforts that lend themselves to this kind of iteration: conversion funnels, performance ad campaigns, notifications are common here.
The “miles” efforts are bigger bets. They break through local maxima and stagnation. They involve a lot more upfront research, data and intuition and most of all, investment in resources. They often take months if not years to build. They may not work but when they do, they can unlock a step-level change in growth. They break existing patterns and may open up new avenues for how your customers use your product. Almost all hardware products have to be definitionally built this way.
The key lesson I’ve learned is: neither model is necessarily better. I’ve seen organizations and products that have been world class at both. You almost always need a portfolio of both to give oxygen to the current product/business while you build out new capabilities. The key is to find the best model to match the product/market/team you have. If you have a mismatch between the team and the model or the product constraints, you’ll almost surely end up with a bad mess on your hands.